PETERS, — METABOLISM AND DIVISION IN PROTOZOA. 491 



of the four salts experimented with give results obtained with animals 

 mostly from one culture, No. 20318, and that these experiments were 

 made within a short period of time. These Stentors originated from a 

 good culture under conditions as nearly representing a pliysiological 

 norm as can well be obtained. The particular results are therefore tlior- 

 oughly comparable among themselves. But comparison withiu the limits 

 of what is jiractically a single experiment does not show the extent of 

 their applicability. When a certain relation between the effects of given 

 substances has been satisfactorily established within the limits of an 

 experiment, that relation can be transferred logically to other experi- 

 ments and will be found to be applicable ; but it may be pitched upon a 

 relatively higher or lower scale according to the metabolic condition of 

 the animals. In other words, nutritional variations within certain limits 

 do not nullify fundamental adjustments of the organism, as one might be 

 led to suppose by a comparison of numerical values obtained at different 

 times. Protozoa are not more capricious in reaction than other groups, 

 their reputation for this resting upon a failure to investigate conditions. 

 There exists for each species a normal physiological condition which the 

 experimenter must learn to recognize. There also exist within the cell, 

 as proved by characteristic reactions, normal substance-relations that are 

 identifiable in series of experiments even when these are conducted at 

 different pitches of metabolic activity. For the detection of these, abso- 

 lute numerical values are no guide. But the relation of values to each 

 other, and a common standard of comparison, enable us to compare 

 experiments differing widely in numerical results. The possibility of 

 doing this is an important consideration when experiments necessarily 

 extend over a long period of time. Concrete illustrations of this method 

 will be given in this and the following sections. 



We may now put the question, AVhat adjustments, that is, what corre- 

 lations of conditions, in Stentor are shown by the preceding data ? Let 

 Hs begin with the curves (p. 473), An unmistakable feature is their 

 serial arrangement. This is shown by the fact that the rise in the value 

 of the mean result above the zero point occurs serially in the order of 

 diminishing concentration as follows: calcic chloride, potassic chloride, 

 sodic chloride, magnesic chloride. Practicall}'^ the same order holds for 

 the opposite ends of the curves (or the curves produced), for the points 

 at which the mean results reach the value of unity, as is graphically 

 represented by their intersections with the axis of abscissas. More or less 

 parallelism is evident in the curves, and no intersection with each other 

 occurs on the paths of the curves between the zero points and the axis 



